Re: Ligurian

--- In cybalist@yahoogroups.com, Bhrihskwobhloukstroy <bhrihstlobhrouzghdhroy@...> wrote:

>
> 2012/5/4, dgkilday57 <dgkilday57@...>:
> >
> > [...]
> >
> > [DGK:]
> > No, my theory explains *bartis as a toponomastic loanword from Ligurian to
> > Celtic, with no phonological criteria applied by the borrowers. Its
> > retention is, pardon the root, fortuitous. Kilday finally gets some good
> > luck!
>
> Bhrihskwobhloukstroy:
> So, where do Yo[u] suggest Ir. bairt comes from? To my limited
> knowledge, there's no place-name whose stem is simple *Barti-s

We should give the Celticists here some time to comment on this hapax legomenon.

> >> If You really think that all these surely plausible but surely ad
> >> hoc conjectures are better than a straightforward Celtic
> >> Lautgesetzlichkeit, please continue, so that all Members will judge by
> >> themselves who is right
>
> > [DGK:]
> > The jury (Piotr, Anders, etc.) is still out on the sequence of soundlaws
> > required for you to have your /o:/-grade.
>
> Bhrihskwobhloukstroy:
> If it comes out that a sound-law has to be revised or changed, I
> immediately modify every reconstruction of mine that can be involved.
> This is obvious and natural.
> As for my theory, I don't think anyone cares about it. I've
> mentioned it for the sake of the argument.
> Maybe this particular question runs the risk of remaining
> unsolved, because one of the two possibilities, i.e. */o:/ > /o/ in
> Osthoff's contexts, would be formally always interpretable as PIE
> short */o/; all the demonstration insists on the scanty evidence of
> not otherwise explainable /a/ (therefore < */a:/ < */o:/), but if on
> one side You propose that every instance is a Ligurian loan (if indeed
> for Ir. hapax bairt, then You'll have no limits for other Celtic
> items: they'll be even closer to Antiquity in time and to Liguria in
> space) and You seriously think that while one Irish attestation
> (bairt, bairte) is uncertain, no attestation at all - in a language
> whose reconstruction is much more hypothetic and moreover not in its
> proper area - is more valid (an attestation in a Celtic derivative
> isn't a proper attestation, because the matter of discussion is
> precisely the original non-Celticity of 'Ligurian' *barti-), then I
> conclude that we don't have any scientific criterion in common (I
> confess that if were in Your situation, I'd have given up much
> earlier, but evidently You have another perception, it's Your full
> right, but the consequence is that we are losing our time).

I am not suggesting Ligurian loanwords in Irish, but there is generally more than one way to skin a hapax. At any rate, there is no reason to give up on a theory which simplifies the interpretation of place-names and regional words in Greater Liguria.

> > [DGK:]
> > In the meantime, you have dumped
> > a truckload of place-names which you have uniformly explained as Celtic,
> > often by merely citing similar-sounding Irish, Welsh, or Breton words.
>
> Bhrihskwobhloukstroy:
>
> My "truckload" is of 5000 characters. I've excerpted it from a
> paper of 110,000 characters, without bibliography (restricted version
> of a three-time longer paper); 110,000 - 5000 = 105,000 characters of
> Celtic and PIE reconstructions. How would You have labeled such a
> greater truckload?

A trainload?

> > [DGK:]
> > Now,
> > I have no doubt that many of these names are indeed Celtic, such as those
> > implying *Brig-.
>
> Bhrihskwobhloukstroy:
>
> 5 km SW of Barzago, whose name You analyse as Ligurian stem +
> Gaulish suffix, there's Briosco < *Brig-usko-, precisely with *Brig-;
> down there, the Celts would have coined a hybrid place-name with
> Celtic root and Ligurian suffix or Celticized its root and retained
> its suffix, just the opposite of *Barti-a:kon. Do You really maintain
> that?

Place-name suffixes can certainly be borrowed from substrate. In the Danelaw, Danish <-by> 'village' is found with English stems, e.g. Willoughby. Thus there can be no a-priori objection to a Gaulish stem taking a Ligurian suffix where Ligurian was spoken.

Earlier I overlooked the possibility of *brig- arising from Gaulish or Ligurian *wrig-. Valtelline <briánz> 'wormwood' points to a protoform *brigantios, disagreeing with Gallo-Latin <bricumus> 'id.' but agreeing with <brigantes> 'parasitic worms of the eyelids' (both from Marcellus of Bordeaux, ca. 400 CE), Welsh <gwraint> 'worms in the skin', Irish <frigh(id)> 'worms in meat', and Breton <grec'h> 'worm'. Worms wriggle, so <brigantes> could continue an obsolete (in Celtic) participial formation, with zero-grade present (like the Sanskrit 6th pres. class) from the root *wreig(^)H- reflected in Germanic *wri:g- (Old English <wri:gian> 'to turn aside, twist', Middle Low German <wriggeln> 'to wriggle'). But <bricumus> appears to reflect instead the zero-grade of *wreik(^)-, whose /o/-grade appears in Gaul. (and Lig.?) *wroikos 'heather' (cf. Old Ir. <froech>, We. <grûg>), whence Gallo-Latin <bru:cus> (glossed once, and reflected in Romance from Catalan to Milanese). I do not know whether *wrikomos 'wormwood' (against *wrigantios) was endemic to SW Gaul; if so, Valt. <briánz> could be Gaulish as well as Ligurian. For 'heather' (also 'broom-plant' and 'Alpine rose'), Venetian <brika>, Fassatalish <breg>, and Piemontese <brek> point to an Illyrian *wraikos, but to the south Sicel(?) *wroikos appears as Calabrian <bruco>, <bruca> and Sicilian <bruca>, <vruca>, <viruca> 'Tamarix gallica'.

> > [DGK:]But your methodology provides no means of distinguishing
> > Celtic from non-Celtic,
>
> Bhrihskwobhloukstroy:
> As everybody knows, none of us is the Voice of Truth. Our task is
> to apply a methodology.
> The methodology I apply is: I chose an area where the presence of a
> given language is assured (in this case, by inscriptions), I apply
> independently verified sound-laws and see what comes out.
> Distinguishing Celtic from non-Celtic is a procedure that can be
> started only *after* that one has completed both the application of
> the Celtic analysis and the application of every non-Celtic analysis.

How silly. If a few place-names suggest that Celtic analysis is inadequate (and they do), there is no point in mechanically constructing a huge corpus of "regelrecht" Celtic etymologies, like a Neogrammarian in the wrong century.

> Dubious and ambiguous cases can emerge only from such a preliminary
> work.
> I'm doing my part for Celtic; You are doing Your part for a
> language that maybe never existed as such, but has been anyway
> seriously proposed and therefore must be taken into consideration.
> Beside that, we have to look for ambiguous Latin/Celtic and
> Germanic/Celtic cases. As Tavi will promptly remember to us, we have
> to take into consideration Basque as well (although no Basque
> linguistic text has ever been found in this area).
> I fear that in [20]57 we won't have completed our tasks yet.
>
> > [DGK:]
> > presuming you have already weeded out Roman and
> > later names.
>
> Bhrihskwobhloukstroy:
> Roman names have been investigated by generations of Italian
> philologists, fortunately so influenced (up to the present day! They
> never realize that) by nationalist prejudice that they have
> marvellously applied the procedure I've just mentioned: to analyze
> everything through Latin / Romance.
> A similar approach was about to be started before II WW in
> Germanic (Gothic and Langobardic) perspective, but some problems, You
> know, have interrupted it... I've tried from time to time to resume
> it, but it's less productive than Celtic, so for the time being I
> privilege the Celtic approach. (On the contrary, the Germanic approach
> is extremely productive in both Rhaeto-Cisalpine and Italian
> surnames.)
> By now we are provisionally in condition to contrast Latin /
> Romance and PIE > Celtic etymologies.
> An example:
> Three km North of Barzio (DGK: the pure Ligurian form without
> -a:ko-?) there's Primaluna, usually understood as Latin pri:ma lu:na
> 'first moon'; I rather compare Welsh 1 pryf, Ir. crum(h) 'worm, larve,
> maggot, fly, insect, small animal of the forest, reptile, snake,
> dragon etc.', and Ir. (con-)lón, (con-)lúan 'moor', Bret. louan
> 'copse': PIE *kwrimo h2lounah2 > *kwrimo:louna: > PC *kwrima:louna: >
> Gaulish *prima:louna: > Latin *Prima:louna > *Prima:lu:na, with
> laryngeal lengthening in composition, Celtic */o:/ > /a:/ and p-Celtic
> */kw/ > /p/.
> There's also some instance of Latin vs. Germanic vs. Celtic
> etymology, but now I have absolutely no more time

Gaulish determinative compounds took the accent on the last syllable of the first element, as shown by certain tribal names which did not shift their accent to conform to the Latin penultimate law: Eburóvi:ces > Évreux, Durócasses > Dreux, Catúri:ges > Chorges, etc. (cf. Dottin, Lang. Gaul. 104). Your hypothetical Gaulish compound *Prima:louna: should thus have had the accent on the antepenult, leading to weakening of the penult, and ultimately to loss of the ultima, in your part of the Romance world. Compare <Albíggaunon> (Strabo), Medieval Latin <in Albingano> (9th cent.), modern <Albénga> (dial. pron. [arbé.Nga] acc. to Petracco Sicardi). Moreover, lowering of pretonic preconsonantal short */i/ occurs in modern <Bedonia> (acc. sg. <Bituniam>, acc. pl. <Bitunias>, Sent. Minuc.), modern <Bresello> against <Brixellum>, and the like. Therefore, I would expect your Gaul. cpd. *Primá:louna: to yield early medieval *Primálona, modern *Premálo or *Premála (perhaps metathesized to *Perm-), but certainly not <Primaluna>.

Nevertheless I think the second half of your etymology can be salvaged. I have never believed the folk-explanation that the coastal town <Lu:na> was named after its alleged crescent-shaped harbor. I would rather posit a Ligurian term *louna: cognate with Breton <louan> 'copse'. This would have yielded an appellative *lu:na 'copse' vel sim. in the local Latin, continuing into the medieval vernacular. *Prima Luna would then simply mean 'Prime Copse' (for cutting wood), 'Beautiful Copse' or the like (cf. Old French <prin>, <prime> 'first-rate, beautiful, delicate' from Lat. <pri:mus>). There is, of course, a world of difference between this straightforward medieval etymology and your highly contrived Gaulish compound, which you seem willing to project all the way back to PIE.

So, while I reject both the "pure" Latin/Romance etymology and your "pure" Celtic etymology, I do not reject your results as "pure" baloney. I find some of them usable (and thanks for Barzio).

> > [DGK:]
> > It is like a broken sieve of Eratosthenes which declares every
> > odd integer a prime. Rather than attempting to deal with the entire
> > truckload at once, I intend to post comments addressing individual
> > place-names as this thread continues.
> >
> > DGK
>
> Bhrihskwobhloukstroy:
>
> Very good! Thank You.
> I immediately dare to propose a case.
> Between Barzio and Primaluna there's Pasturo, cf. Oir. 1 cas
> 'curly, intricate', Middle Ir. 2 úr 'green': PIE *kwösto-puh2ro- >
> *kwösto-pu:ro- > PC *kwasto-[p]u:ro- > Gaulish *pastou:ro- > Latin
> *Pastouru- > *Pastu:ru-.
> Do You have objections?

Are you beginning to catch a glimmer of the trouble with your
methodology? You preconclude that a place-name where Celts once lived must have a Celtic etymology, so you thumb through your dictionaries and invent one: <Pasturo>, you insist, means 'Intricate Green' (an excellent site for a challenging golf course). A "pure" Latino-Romanist could equally well argue that Latin-speakers once lived there, and preconclude that <Pasturo> has an obvious Latin etymology, 'Fattened Aurochs' (the site of an archaic Prodigal Son reception, perhaps).

If we are to get serious about this place-name, I must ask you for some information, since you are geographically much closer to the action than I am:

1. What is the local pronunciation of <Pasturo>, including the accent? (You can use Petracco Sicardi's notation or some convenient adaptation.) Are there significantly different variants in other villages?

2. What are the earliest written attestations of the name? How do these change through the centuries?

3. Might the name have undergone folk-etymological influence? Are there phonetically close words in the vernacular, the standard administrative language, or learned language, such as reflexes of LL <pa:stu:ra> 'pasture', <pa:sto:ria> 'shepherd's hobble', etc.?

DGK